sigmashake-docs — Paging policy
Who responds to a problem with the docs site, and how. Rotation in oncall.toml.
This policy is best-effort. No paging vendor is connected. No 24/7 coverage is promised. No acknowledgement time and no resolve time is promised. One role, the ISMS owner, responds.
Note 2026-10-11: an earlier version of this page described PagerDuty paging with a 10-minute, 24/7 acknowledgement. No PagerDuty integration exists. That text is removed.
Severity matrix
| Severity | Definition | Response |
|---|---|---|
| P1 | Customer-visible outage (site unreachable or 5xx for all users, security incident, data-loss event) | The ISMS owner triages under the incident response plan (triage target 1 hour, best effort) |
| P2 | Customer-visible degradation (slow pages, one region failing) | The ISMS owner reviews in normal working time |
| P3 | Trend or warning (latency trend, dependency CVE high) | The ISMS owner reviews in normal working time |
| P4 | Informational (no deploy-completed or weekly SLO notice is generated today) | No action |
What counts as P1
- Sustained 5xx responses for the whole site
- Any data-loss event recorded
- Any security incident
What does NOT count
- Single 5xx response
- Cron job duration longer than expected (the sweep is allowed to take longer)
- Cron job missed at the expected time (Cloudflare can shift a cron by about 15 minutes)
After-hours rules
No after-hours coverage is promised. Work on a P1 starts when the ISMS owner sees the report.
Escalation
The ISMS owner is the only role that responds today. No escalation path and no
automatic re-page exist. The escalation = "self" setting in oncall.toml
means the same role handles every level.